Data overview

## Rows: 384
## Columns: 4
## $ assembly                   <fct> WA, WA, WA, WA, WA, WA, WA, WA, WA, WA, WA,…
## $ revenue_item               <fct> Property Rates, Property Rates, Property Ra…
## $ year                       <chr> "2017", "2018", "2019", "2020", "2021", "20…
## $ diff_btw_actual_and_budget <dbl> NA, 0.00, 178605.70, 218668.70, 35906.50, 5…
##           assembly                     revenue_item     year          
##  BOLGATANGA   :64   Fees                     :48    Length:384        
##  EAST MAMPRUSI:64   Fines                    :48    Class :character  
##  TAMALE       :64   Investment               :48    Mode  :character  
##  WA           :64   Land (incl Miscellaneous):48                      
##  WEST GONJA   :64   Licences                 :48                      
##  YENDI        :64   Other Rates              :48                      
##                     (Other)                  :96                      
##  diff_btw_actual_and_budget
##  Min.   :-419112           
##  1st Qu.:   -564           
##  Median :   2950           
##  Mean   :  27760           
##  3rd Qu.:  40749           
##  Max.   : 555674           
##  NA's   :43
## # A tibble: 10 × 4
##    assembly revenue_item   year  diff_btw_actual_and_budget
##    <fct>    <fct>          <chr>                      <dbl>
##  1 WA       Property Rates 2017                         NA 
##  2 WA       Property Rates 2018                          0 
##  3 WA       Property Rates 2019                     178606.
##  4 WA       Property Rates 2020                     218669.
##  5 WA       Property Rates 2021                      35906.
##  6 WA       Property Rates 2022                       5493.
##  7 WA       Property Rates 2023                      41017.
##  8 WA       Property Rates 2024                         NA 
##  9 WA       Other Rates    2017                         NA 
## 10 WA       Other Rates    2018                      89794
## [1] Property Rates            Other Rates              
## [3] Fees                      Fines                    
## [5] Licences                  Land (incl Miscellaneous)
## [7] Rent                      Investment               
## 8 Levels: Fees Fines Investment Land (incl Miscellaneous) ... Rent
## [1] "2017" "2018" "2019" "2020" "2021" "2022" "2023" "2024"
## [1] WA            TAMALE        BOLGATANGA    YENDI         EAST MAMPRUSI
## [6] WEST GONJA   
## Levels: BOLGATANGA EAST MAMPRUSI TAMALE WA WEST GONJA YENDI
## --- Distribution of Observations ---
Assembly Distribution
assembly n Proportion
BOLGATANGA 64 0.167
EAST MAMPRUSI 64 0.167
TAMALE 64 0.167
WA 64 0.167
WEST GONJA 64 0.167
YENDI 64 0.167
Revenue Item Distribution
revenue_item n Proportion
Fees 48 0.125
Fines 48 0.125
Investment 48 0.125
Land (incl Miscellaneous) 48 0.125
Licences 48 0.125
Other Rates 48 0.125
Property Rates 48 0.125
Rent 48 0.125
Year Distribution
year n Proportion
2017 48 0.125
2018 48 0.125
2019 48 0.125
2020 48 0.125
2021 48 0.125
2022 48 0.125
2023 48 0.125
2024 48 0.125
Detailed Summary of Difference Between Actual and Budget
skim_variable numeric.mean numeric.sd numeric.p0 numeric.p25 numeric.p50 numeric.p75 numeric.p100 numeric.hist
data 27759.5 114126.2 -419112.2 -564 2950 40749 555673.8 ▁▁▇▁▁
Skewness and Kurtosis of Difference Between Actual and Budget
Skewness Kurtosis
1.62 10.11

The dataset (384 rows, 4 columns: assembly, revenue_item, year, diff_btw_actual_and_budget), captures the difference between actual and budgeted revenue (in unspecified monetary units) across 6 assemblies (WA, Tamale, Bolgatanga, Yendi, East Mamprusi, West Gonja), 8 revenue items (Property Rates, Other Rates, Fees, Fines, Licences, Land (incl Miscellaneous), Rent, Investment), and years 2017–2024. Each assembly (64 rows, 16.7%), revenue item (48 rows, 12.5%), and year (48 rows, 12.5%) is equally represented.

The diff_btw_actual_and_budget is right-skewed (mean = 27,760, median = 2,950, min = -419,112, max = 555,674). (skewness = 1.62, kurtosis = 10.11)

The range spans -419,112 (underperformance) to 555,674 (overperformance), with the 1st quartile at -564 and 3rd quartile at 40,749. The histogram below confirms a peak near zero and a long right tail.

Most differences are small (median = 2,950), but rare large overperformances ( 555,674) inflate the mean.

Descriptive

Difference by Assembly
assembly Mean Median SD Min Max N
BOLGATANGA 33129.09 0.0 111451.58 -263989.0 495725.6 64
EAST MAMPRUSI 3051.79 1014.5 10910.08 -20535.0 68000.0 64
TAMALE 56920.93 8543.1 160411.06 -300950.3 550546.4 64
WA 21080.58 0.0 168445.32 -419112.2 555673.8 64
WEST GONJA 22424.33 3600.0 85670.78 -229724.0 300000.0 64
YENDI 29255.69 10919.0 64003.61 -100000.0 287169.0 64
Difference by Revenue Item
revenue_item Mean Median SD Min Max N
Fees 29526.59 11993.5 144852.29 -317392.10 555673.8 48
Fines 14887.59 0.0 69857.86 -54174.27 430829.7 48
Investment -1172.14 0.0 56614.85 -229724.00 112399.2 48
Land (incl Miscellaneous) 52740.71 18121.7 125543.93 -102997.00 508061.9 48
Licences 14302.73 2934.3 136790.97 -419112.20 373412.8 48
Other Rates 5681.11 1029.0 61819.55 -300950.30 157861.0 48
Property Rates 56938.21 9443.8 136433.59 -259436.00 550546.4 48
Rent 48822.86 5646.0 128265.56 -175160.00 465747.0 48
Difference by Year
year Mean Median SD Min Max N
2017 21804.77 3100.00 190308.55 -263989.0 385643.6 48
2018 19479.02 3600.00 74341.94 -89552.0 287169.0 48
2019 5110.58 0.00 81955.25 -229724.0 300000.0 48
2020 27653.50 4300.00 103197.92 -205977.9 555673.8 48
2021 44246.39 4205.50 144005.76 -419112.2 508061.9 48
2022 30684.87 2649.25 113770.52 -259436.0 465747.0 48
2023 26711.12 725.00 139981.44 -300950.3 550546.4 48
2024 41411.33 9285.59 112899.37 -210947.0 522288.4 48
assembly avg_diff deficit_pct largest_surplus largest_deficit volatility
TAMALE 56921 23 550546 -300950 160411
BOLGATANGA 33129 32 495726 -263989 111452
YENDI 29256 21 287169 -100000 64004
WEST GONJA 22424 23 300000 -229724 85671
WA 21081 35 555674 -419112 168445
EAST MAMPRUSI 3052 20 68000 -20535 10910
revenue_item total_diff avg_abs_diff surplus_ratio
Property Rates 2391405 79238 1
Land (incl Miscellaneous) 2267851 83295 1
Rent 2099383 84463 1
Fees 1269644 77293 1
Fines 640166 22439 0
Licences 615018 86331 1
Other Rates 232926 25901 1
Investment -50402 22025 0
year net_diff avg_diff deficit_count
2017 152633 21805 2
2018 915514 19479 13
2019 245308 5111 17
2020 1327368 27654 13
2021 2123827 44246 10
2022 1472874 30685 10
2023 1282134 26711 16
2024 1946332 41411 7

Assembly Performance: Tamale Leads, East Mamprusi Lags:

Means: Tamale has the highest mean difference (56,921, max = 550,546), followed by Bolgatanga (33,129, max = 495,726), Yendi (29,256), West Gonja (22,424), WA (21,081, max = 555,674), and East Mamprusi (3,052, max = 68,000).

Medians: Yendi (10,919) and Tamale (8,543) lead, while Bolgatanga and WA have medians of 0, indicating frequent on-target or balanced performance.

Volatility: WA (SD = 168,445) and Tamale (SD = 160,411) are most volatile, with large deficits (e.g., WA’s -419,112). East Mamprusi is least volatile (SD = 10,910).

Deficit Rates: WA has the highest deficit percentage (35%), Bolgatanga 32%, while East Mamprusi is lowest (20%). Surplus ratios are high across assemblies (65–80%).

Implication: Tamale’s strong performance aligns with contribution_to_rev’s high median (9.95%) and rev_growth’s leadership (mean = 2,047.67%). East Mamprusi’s low differences mirror its contribution_to_rev underperformance (median = 6.35%).

Revenue Item Performance: Property Rates and Land Dominate:

Means: Property Rates (mean = 56,938, median = 9,444) and Land (52,741, median = 18,122) lead, followed by Rent (48,823, median = 5,646) and Fees (29,527, median = 11,994). Investment is negative (mean = -1,172, median = 0).

Volatility: Fees (SD = 144,852), Licences (136,791), and Property Rates (136,434) are most volatile, with large deficits (Licences’ -419,112). Fines (SD = 69,858) and Investment (56,615) are less volatile.

Surplus Ratios: Property Rates, Land, Rent, Fees, Licences, and Other Rates have 100% surplus ratios (all non-NA differences positive), while Fines and Investment have 0% (all deficits or zero).

Total Differences: Property Rates (2,391,405) and Land (2,267,851) contribute most to surpluses, while Investment (-50,402) drives deficits.

Implication: Property Rates’ dominance contrasts with contribution_to_rev’s Fees/Licences leadership (medians 21.9%, 35%), suggesting budgeting accuracy issues for Fees/Licences. Investment’s deficits align with contribution_to_rev’s weakness (median = 0%).

Temporal Trends: 2021–2024 Peaks, 2019 Low:

Means: Highest in 2021 (44,246), 2024 (41,411), and 2022 (30,685); lowest in 2019 (5,111). Medians peak in 2024 (9,286), with 2019 at 0.

Volatility: 2017 is most volatile (SD = 190,309), followed by 2021 (144,006) and 2023 (139,981). 2018 is least volatile (SD = 74,342).

Net Differences: Highest in 2021 (2,123,827) and 2024 (1,946,332); lowest in 2017 (152,633). Deficit counts peak in 2019 (17) and 2023 (16), lowest in 2017 (2) and 2024 (7).

Implication: The 2021–2024 surplus peaks align with contribution_to_rev’s 2020–2021 highs (medians 10.35–12.05%) and rev_growth’s 2023 spike (mean = 1,479.85%). The 2019 low suggests budgeting challenges, possibly tied to policy shifts.

Line Plots (Median): Yendi (10,919) and Tamale (8,543) lead in medians, with Bolgatanga/WA at 0, indicating frequent on-target performance. Medians peak in 2024 (9,286) and dip in 2019 (0).

Box Plots: Tamale and WA have wide ranges (SDs 160,411, 168,445) with (Tamale’s 550,546, WA’s -419,112). East Mamprusi’s tight range (SD = 10,910) reflects low variability.

Heatmap (Assembly-Year): Tamale’s surpluses peak in 2021–2024 ( 550,546 in 2023), WA shows deficits in 2021 (-419,112). Bolgatanga and West Gonja have moderate surpluses.

Bar Plot (Total Difference): Tamale contributes the highest total surplus (3,641,340), followed by Bolgatanga (2,120,260). East Mamprusi’s total is minimal (~195,360).

Comparison: Tamale’s strength aligns with contribution_to_rev’s high median (9.95%) and rev_growth’s leadership (mean = 2,047.67). East Mamprusi’s low performance mirrors its contribution_to_rev underperformance (median = 6.35%).

Implication: Tamale’s budgeting success offers a model for Bolgatanga/East Mamprusi, where deficits and low surpluses indicate forecasting weaknesses.

Revenue Item Trends: Property Rates and Land Lead, Investment Lags

Line Plots (Mean): Property Rates (mean = 56,938) and Land (52,741) show the highest mean differences, peaking in 2021–2024. Fees (29,527) and Rent (48,823) are moderate, while Investment (-1,172) and Fines (14,888) are low or negative.

Line Plots (Median): Land (18,122), Fees (11,994), and Property Rates (9,444) have high medians, while Fines and Investment are 0, reflecting frequent deficits or on-target performance.

Box Plots: Property Rates, Land, and Fees have wide ranges (SDs > 125,000) with outliers (e.g., Property Rates’ 550,546). Investment and Fines have tight ranges (SDs < 70,000), with medians at 0.

Bar Plot (Total Difference): Property Rates (2,391,405) and Land (2,267,851) contribute the most to surpluses, followed by Rent (2,099,383). Investment (-50,402) is the only net deficit.

Histogram (Performance): Property Rates, Land, Rent, Fees, Licences, and Other Rates show 100% surplus ratios (all non-NA differences positive). Fines and Investment have 0% surplus ratios, indicating deficits.

Alluvial Plot: Property Rates and Land increase in share of total differences (2021–2024), while Fines and Investment decline, reflecting budgeting shifts.

Comparison: Property Rates’ dominance here contrasts with contribution_to_rev’s Fees/Licences leadership (medians 21.9%, 35%). Investment’s deficits align with contribution_to_rev’s weakness (median = 0%) and rev_growth’s stagnation (median = 0%).

Implication: Property Rates and Land are reliable for surpluses, but Fees/Licences’ moderate performance suggests budgeting inaccuracies. Investment requires reform.

Temporal Trends: 2021–2024 Surpluses, 2019 Low

Line Plots (Mean): Mean differences peak in 2021 (44,246), 2024 (41,411), and 2022 (30,685), with a low in 2019 (5,111). 2017–2018 show moderate surpluses (19,479–21,805).

Line Plots (Median): Medians peak in 2024 (9,286), followed by 2020–2021 (~4,300), with a low in 2019 (0). 2023’s low median (725) contrasts with its high mean (26,711).

Box Plots (Year): 2021 and 2024 have wide ranges (SDs 144,006, 112,899) with outliers ( 555,674 in 2020). 2019 has the tightest range (SD = 81,955).

Heatmap (Assembly-Year): 2021–2024 show strong surpluses ( Tamale in 2023), while 2019 has deficits across assemblies ( West Gonja’s -229,724).

Line Plot (Cumulative Deviations): Cumulative surpluses peak in 2021 (2,123,827) and 2024 (1,946,332), with 2019 lowest (245,308). Deficit counts are highest in 2019 (17) and 2023 (16), lowest in 2017 (2) and 2024 (7).

Comparison: The 2021–2024 peaks align with contribution_to_rev’s 2020–2021 highs (medians 10.35–12.05%) and rev_growth’s 2023 spike (mean = 1,479.85%). The 2019 low suggests a challenging year, possibly due to policy or economic factors.

Implication: The 2021–2024 surplus trend indicates effective collection, but 2019’s deficits highlight forecasting issues to address for 2025.

Warnings: NA-related row removals ( 43 in histogram, 5 in median line plots)

Policy and Strategic Implications

Budget Planning: Leverage Property Rates and Land for 2025 surplus forecasts, but improve Fees/Licences budgeting accuracy (medians 11,994, 2,934 vs. contribution_to_rev’s 21.9%, 35%). Anticipate 2024-like surpluses (median = 9,286).

Regional Focus: Replicate Tamale’s success (mean = 56,921) in Bolgatanga/East Mamprusi, where low surpluses (3,052–33,129) and high deficits ( WA’s 35%) persist.

Item Reforms: Address Investment’s deficits (-50,402) and Fines’ low surpluses (0% ratio) through policy changes ( investment incentives), aligning with contribution_to_rev’s findings.

Kruskal-Wallis Test: Difference by Assembly
Test Chi_Squared P_Value
Kruskal-Wallis (Assembly) 8.227 0.144
Kruskal-Wallis Test: Difference by Revenue Item
Test Chi_Squared P_Value
Kruskal-Wallis (Revenue Item) 16.241 0.023

The Kruskal-Wallis test for assemblies yields a chi-squared statistic of 8.227 with a p-value of 0.144 (> 0.05), indicating no statistically significant differences in diff_btw_actual_and_budget across the 6 assemblies (WA, Tamale, Bolgatanga, Yendi, East Mamprusi, West Gonja).

Context from Prior Analysis: Despite Tamale’s high mean difference (56,921, median = 8,543) and East Mamprusi’s low mean (3,052, median = 1,014), the high variability (SDs up to 168,445 for WA) and skewness (1.62) likely obscure statistical differences. Visualizations ( box plots) showed Tamale and WA with wider ranges and outliers Tamale’s 550,546, WA’s -419,112), but medians for Bolgatanga/WA were 0, aligning with frequent on-target performance.

Implication: While statistical significance is absent, practical differences (for ex. Tamale’s surpluses vs. East Mamprusi’s low performance) suggest targeted budgeting interventions for weaker assemblies like Bolgatanga and East Mamprusi, which showed high deficit rates (32% and 20%, respectively).

The Kruskal-Wallis test for revenue items yields a chi-squared statistic of 16.241 with a p-value of 0.023 (< 0.05), indicating statistically significant differences in diff_btw_actual_and_budget across the 8 revenue items (Property Rates, Land (incl Miscellaneous), Rent, Fees, Licences, Other Rates, Fines, Investment).

From Prior Analysis: Property Rates (mean = 56,938, median = 9,444) and Land (52,741, median = 18,122) led in surpluses, with total differences of 2,391,405 and 2,267,851, respectively. Investment (mean = -1,172, median = 0) and Fines (14,888, median = 0) had low or negative contributions, with Investment showing a net deficit (-50,402). Box plots and line plots confirmed Property Rates/Land’s high medians and volatility (SDs > 125,000), while Fines/Investment had tight ranges (SDs < 70,000). The alluvial plot showed Property Rates/Land increasing in share (2021–2024), while Fines/Investment declined.

Conclusion

The diff_btw_actual_and_budget exhibits a right-skewed distribution (skewness = 1.62, kurtosis = 10.11), with a median of 2,950 and say outliers driving the mean to 27,760 (range: -419,112 to 555,674).

Implication: The median is a robust budgeting benchmark.

Assembly Performance Varies Practically but Not Statistically:

Tamale leads with the highest mean (56,921) and median (8,543) differences, followed by Bolgatanga (33,129, median = 0) and Yendi (29,256, median = 10,919). East Mamprusi lags (mean = 3,052, median = 1,014). WA and Tamale are most volatile (SDs 168,445, 160,411), with big deficits ( WA’s -419,112) and surpluses (Tamale’s 550,546). Deficit rates are highest in WA (35%) and Bolgatanga (32%).

Kruskal-Wallis Test: No significant differences across assemblies (chi-squared = 8.227, p = 0.144), likely due to skewness and outliers, as seen in contribution_to_rev (p = 0.923).

Visualizations: Line plots and heatmaps show Tamale’s consistent surpluses (2021–2024), while box plots highlight WA/Tamale’s outliers. Bar plots confirm Tamale’s highest total surplus (3,641,340).

Comparison: Tamale’s strength mirrors contribution_to_rev’s high median (9.95%) and rev_growth’s leadership (mean = 2,047.67). East Mamprusi’s underperformance aligns with its low contribution_to_rev median (6.35%).

Implication: Despite statistical non-significance, Tamale’s budgeting success offers a model for Bolgatanga and East Mamprusi, where deficits and low surpluses indicate forecasting weaknesses.

Revenue Items Show Significant Differences, with Property Rates and Land Leading:

Property Rates (mean = 56,938, median = 9,444) and Land (52,741, median = 18,122) drive surpluses (total differences: 2,391,405, 2,267,851), followed by Rent (48,823, median = 5,646) and Fees (29,527, median = 11,994). Investment (-1,172, median = 0) and Fines (14,888, median = 0) contribute deficits or low surpluses, with Investment at -50,402.

Kruskal-Wallis Test: Significant differences across revenue items (chi-squared = 16.241, p = 0.023), reflecting Property Rates/Land’s outperformance vs. Investment/Fines’ weakness.

Visualizations: Box plots show Property Rates/Land’s high medians and volatility (SDs > 125,000), while Fines/Investment have tight ranges (SDs < 70,000). Alluvial plots indicate Property Rates/Land’s increasing share (2021–2024), with Fines/Investment declining. Histograms confirm 100% surplus ratios for most items, except Fines/Investment (0%).

Comparison: Property Rates’ dominance contrasts with contribution_to_rev’s Fees/Licences leadership (medians 21.9%, 35%). Investment/Fines’ deficits align with contribution_to_rev’s low medians (0.3%, 0%) and rev_growth’s stagnation (median = 0%).

Implication: Property Rates and Land are reliable for surplus generation, but Fees/Licences’ moderate budgeting performance suggests forecasting inaccuracies. Investment/Fines require structural reforms.

Temporal Trends Highlight 2021–2024 Surpluses, 2019 Deficit Low:

Mean differences peak in 2021 (44,246), 2024 (41,411), and 2022 (30,685), with a low in 2019 (5,111). Medians peak in 2024 (9,286), dipping in 2019 (0). Cumulative surpluses are highest in 2021 (2,123,827) and 2024 (1,946,332), lowest in 2017 (152,633). Deficit counts peak in 2019 (17) and 2023 (16), lowest in 2017 (2) and 2024 (7).

Visualizations: Line plots and heatmaps confirm 2021–2024 surplus peaks (e.g., Tamale’s 550,546 in 2023), with 2019 deficits across assemblies ( West Gonja’s -229,724). Box plots show 2021/2024’s wide ranges (SDs > 112,000).

Comparison: The 2021–2024 peaks align with contribution_to_rev’s 2020–2021 highs (medians 10.35–12.05%) and rev_growth’s 2023 spike (mean = 1,479.85). The 2019 low suggests a challenging year, consistent across datasets.

Implication: The 2021–2024 surplus trend indicates effective collection, but 2019’s deficits highlight forecasting vulnerabilities to address for 2025.

Strategic Implications for possible budgeting or budgeting

Optimize Revenue Item Budgeting:

Action: Prioritize Property Rates and Land for surplus forecasts, given their high medians (9,444, 18,122) and significant contributions (total differences > 2.2M). Align Fees/Licences budgeting with their contribution_to_rev dominance (medians 21.9%, 35%) to reduce discrepancies (current medians 11,994, 2,934).

Reform: Overhaul Investment and Fines budgeting, where deficits (-50,402) and low surplus ratios (0%) persist. Introduce incentives or enforcement to boost performance, as seen in contribution_to_rev’s low medians (0.3%, 0%).

Address Regional Disparities: Action: Replicate Tamale’s budgeting success (mean = 56,921, total = 3,641,340) in Bolgatanga (32% deficit rate) and East Mamprusi (mean = 3,052), using Property Rates/Land strategies. Tailor interventions to reduce WA’s high deficits (35%).

Support: Provide technical assistance to weaker assemblies, leveraging Tamale’s 2021–2024 surplus model, akin to contribution_to_rev’s regional gaps.

Build on Temporal Strengths: Action: Capitalize on 2021–2024 surplus trends (peak median = 9,286 in 2024) for 2025 planning, focusing on Property Rates/Land. Investigate 2019’s deficit low (median = 0) to prevent recurrence, possibly due to policy or economic disruptions.

Recommendations

Short-Term (2025 Budget): Set conservative budgets using median differences (2,950), prioritizing Property Rates/Land. Adjust Fees/Licences forecasts upward and reform Investment/Fines. Support Bolgatanga/East Mamprusi with Tamale’s strategies.

Long-Term: Develop policies to enhance Investment/Fines performance ( tax incentives). Standardize data collection to minimize NAs, aligning with contribution_to_rev’s lower NA rate (3.2%).